JRRD Volume 41, Number 5, Pages 695706
September/October 2004
Journal of Rehabilitation Research & Development
Questionnaire for Persons with a Transfemoral Amputation (Q-TFA):
Initial validity and reliability of a new outcome measure
Kerstin Hagberg, PT; Rickard Brnemark, MD, PhD; Olle Hgg, MD, PhD
Department of Prosthetics and Orthotics, Sahlgrenska University Hospital, Gteborg, Sweden; Department
of Orthopaedics, Gteborg University, Gteborg, Sweden
AbstractThe Questionnaire for Persons with a Transfemoral
Amputation (Q-TFA) is a new self-report measure developed
for nonelderly transfemoral amputees using a socket- or
osseointegrated prosthesis to reflect use, mobility, problems,
and global health, each in a separate score (0100). This paper
describes the initial measurement properties of the Q-TFA as
completed by 156 persons with a transfemoral amputation
using a socket prosthesis (67% male, 92% nonvascular cases,
mean age 51 years). Criterion validity was determined by asso-
ciations between scores of the Q-TFA and the Short-Form 36
(SF-36)-Item Health Survey. Reliability was assessed by retest
(n = 48) and by determination of the internal consistency. Cor-
relations between Q-TFA and SF-36-Item Health Survey scales
matched hypothesized patterns. Intraclass correlations were
between 0.89 and 0.97, and measurement error ranged from 10
to 19 points. Cronbachs alpha revealed good internal consis-
tency, with no values less than 0.7. This study shows that the
Q-TFA, applied to persons using a transfemoral socket prosthe-
sis, has adequate initial validity and reliability.
Key words: amputation, artificial limbs, lower limb, osseoin-
tegration, outcome assessment, prosthesis and implants, ques-
tionnaires, reliability and validity, walking.
INTRODUCTION
Lower-limb amputation is a permanent impairment
that has been shown to lead to decreased health-related
quality of life (HRQL), activity limitations, and partici-
pation restrictions [15]. In Sweden, as well as in the rest
of the developed world, the majority of lower-limb
695
amputations are performed because of vascular disease,
with the high mortality rate and average age typical of
this population [68]. The ultimate goal for rehabilitation
after a lower-limb amputation is to achieve the best pos-
sible quality of life for the individual, which in most
cases includes mobility with a prosthetic limb. The suc-
cess of prosthetic rehabilitation is influenced by both the
cause and level of the amputation: vascular cases demon-
strate poorer results than nonvascular cases, and trans-
femoral cases demonstrate poorer results than transtibial
cases [811]. Among all persons using an artificial limb,
socket-related problems and discomfort also constitute
major concerns affecting quality of life and prosthetic
mobility [5,1216].
Abbreviations: HRQL = health-related quality of life, ICC =
intraclass correlation coefficient, MCS = Mental Component
Score, Med = median, PCS = Physical Component Score, PEQ =
Prosthesis Evaluation Questionnaire, Q-TFA = Questionnaire for
Persons with a Transfemoral Amputation, SD = standard devia-
tion, SF-36 = Short-Form 36.
This material was based on work supported by Greta and
Einar Askers Stiftelse and Johan Janssons Stiftelse, all of
Sweden.
Address all correspondence to Kerstin Hagberg, Department of
Orthopaedics, Bruna Strket 11b, Plan 5, Sahlgrenska Univer-
sity Hospital, S-413 45 Gteborg, Sweden; +46-31-343-8133;
fax: +46-31-342-2630; email: kerstin.hagberg@vgregion.se.
DOI: 10.1682/JRRD.2003.11.0167
696
JRRD, Volume 41, Number 5, 2004
A specific subgroup of individuals with amputation
is the group of nonelderly persons having an amputation
for reasons other than vascular disease. This group is a
clear minority in the developed world, but constitutes a
substantial portion of those having an amputation due to
trauma or tumor [17,18]. Those who survive the trauma
or malignancy have a long life expectancy [7].
In our practice, we have a specific need to monitor
changes in status for nonelderly individuals having a
transfemoral amputation to assess those patients who
receive a direct bone-anchored amputation prosthesis
through the method of osseointegration [19]. This
method includes a two-stage surgery in which a titanium
fixture is implanted in the residual bone in the first stage,
and an abutment penetrating the skin is inserted at a sec-
ond surgery. Prosthetic attachment to the residual limb is
obtained by connection of the prosthetic limb directly to
the abutment. Patients most suitable to be treated with a
bone-anchored prosthetic limb have so far been younger
or middle-aged transfemoral amputees, without vascular
disease, facing severe problems in using an artificial limb
suspended with a conventional socket. To optimize the
selection of suitable patients and to study outcomes
related to the change of prosthetic attachment, we need a
useful instrument to reflect changes after treatment with a
bone-anchored prosthesis.
Previously described outcome instruments targeted
for individuals with lower-limb amputation have not been
specifically designed to address the needs of nonelderly
persons having a transfemoral amputation [20,21]. The
Locomotor Capability Index, which is included in the
Prosthetic Profile of the Amputee, has been shown to
have a high ceiling effect and is recommended for use on
individuals with lower prosthetic mobility capabilities
[22]. Therefore, we concluded that development of a new
tool that could be expected to be sensitive for changes in
mobility and problems in use with a prosthesis for our
specific subgroup of nonelderly patients was necessary.
Such a tool could be used to monitor changes in status in
response to, for example, new prosthetic components or a
surgical revision. During the process of development of
our new instrument, the Prosthesis Evaluation Question-
naire (PEQ) was published [23]. This is a valid and reli-
able instrument developed to measure small differences
in prosthesis function and major life domains related to
prosthesis function among persons with a lower-limb
amputation. However, no study has reported on the sensi-
tivity of the PEQ within the subgroup of nonelderly indi-
viduals having a transfemoral amputation, and there is no
valid translation to Swedish.
The patients subjective experience is the issue of
most importance in measurements of health outcomes
[24,25]. Self-report questionnaires are the preferred for-
mat for such measurements [26]. Ideally, such a question-
naire should be clinically relevant, consistent, and brief
and should cover those issues that are considered of par-
ticular interest to the study [27]. Moreover, outcome
questionnaires should yield reliable results if used repeat-
edly on the same patient with a stable condition.
The Questionnaire for persons with a Transfemoral
Amputation (Q-TFA) is a targeted self-report outcome
measure that reflects current prosthetic use, mobility,
problems, and health. Primarily designed for nonelderly
persons having a transfemoral amputation, the Q-TFA
was also developed to study outcomes when changing
from a conventional socket prosthesis to a bone-anchored
prosthesis. Our objective in this study was to assess the
initial validity and reliability of the Q-TFA in individuals
using a transfemoral socket prosthesis in Sweden.
METHODS
Study Population
Subjects were selected based on the following inclu-
sion criteria: adults aged 20 to 70 years, with a unilateral
transfemoral amputation at least 2 years before the study,
using a conventional socket prosthesis, and able to read
and understand the Swedish language. A prosthetic user
was defined as a person who wears a prosthesis at least
once a week, according to Gris et al. [20]. Participants
in the study were recruited in two stages: first, through
national associations for amputees and selected pros-
thetic workshops and/or rehabilitation units, and second,
through all four prosthetic workshops in the county of
Vstra Gtaland.
The Human Research Ethics Committee of the Medi-
cal Faculty at Gteborg University approved the study.
Eligible persons received written information regarding
the study, anonymity, and notification of their right to
discontinue their participation at any time, before signing
the consent form.
Procedures
Two questionnaires (Q-TFA and the Short-Form 36
[SF-36]-Item Health Survey) were sent by mail, together

with questions regarding demographic characteristics and
baseline information. Those who did not answer received
two reminders. Participants living in the county of Vstra
Gtaland were also asked to participate in a test-retest
reliability study: 2 weeks after the questionnaires were
received, we mailed the Q-TFA a second time, along with
four additional questions on important changes of health
condition and condition of the prosthesis to verify a sta-
ble situation since the first mailing. A 2-week interval
between test and retest was chosen because this interval
has been suggested to be long enough to prevent memory
bias but short enough to assure a stable condition in most
cases [28]. Of the 62 persons asked to answer the ques-
tionnaire a second time, 9 were excluded because of
reported changes in condition that were considered clini-
cally significant, and 5 failed to answer, leaving 48 par-
ticipants for this component of the study.
Q-TFA: Development and Description
The aim of the Q-TFA is to determine current status
regarding prosthetic use, prosthetic mobility, problems,
and global health of persons using a transfemoral pros-
thesis. Experts involved in the treatment of such patients,
including two orthopedic surgeons, a prosthetist, and a
physical therapist, identified items selected for construc-
tion of the Q-TFA. Their clinical experience, combined
with a review of existing literature and a continuous dis-
cussion during rehabilitation of patients, formed the
foundation for the questionnaire design. Further, a small
group of patients (n = 4), not included in the study, with
experience using first a conventional socket prosthesis
and then a bone-anchored prosthesis, were consulted in
the development of the instrument.
The Q-TFA is a self-report questionnaire. The cur-
rent version consists of 70 questions, and the time to
complete it is approximately 20 minutes. Based on the
answers to the questionnaire, we constructed a scoring
system. Of the 70 questions, 54 are condensed to four
separate scores: Prosthetic Use Score (Use), Prosthetic
Mobility Score (Mobility), Problem Score (Problem),
and Global Score (Global) (Appendix, available in the
online version only). Each raw score is transformed to a
range from 0 to 100 with the standard scoring method
[27]. The remaining 16 questions, describing details con-
nected with prosthetic use or mobility (i.e., preference to
use the prosthesis or not while cooking, driving, or
socializing, and details of reasons to refrain from using
the prosthetic limb), are not included in any score
697
HAGBERG et al. Questionnaire after transfemoral amputation
because of statistical redundancy or for other reasons.
Those 16 questions are not analyzed in this study and are
not included in the Appendix (online version only).
Prosthetic Use Score (2 Items)
Prosthetic use is defined as the amount of normal
prosthetic wear per week. The number of days per week
the prosthesis is normally worn is multiplied by the num-
ber of hours per day. A Use score of 100 indicates that the
prosthesis normally is worn every day for more than
15 hours a day.
Prosthetic Mobility Score (19 Items)
Prosthetic mobility is defined as the ability and per-
formance of the amputee to move and change and main-
tain postures when using the prosthesis. The score
consists of three subscores, each with a range from 0 to
100: capability (12 items), use of walking aids (2 items),
and walking habits (5 items). The average of these three
subscores generates the total Mobility score. Capability
items consist of questions on the ability to perform loco-
motor activities, independent of the level of difficulty to
perform them. The subscores of walking aids and walking
habits are estimates of prosthetic performance rather than
capability. A result of 100 in the subscore of walking aids
indicates that, in general, the person does not use walking
aids at all. With regard to walking habits, the patient is
asked to answer how often he/she had walked continu-
ously various distances outdoors during the last 3 months.
To summarize, the Mobility score consists of the average
of three subscores (capability, walking aids, and walking
habits). A score of 100 indicates the best possible pros-
thetic mobility as measured with this instrument.
Problem Score (30 Items)
Problems are defined as the extent of specific prob-
lems related to the amputation and the prosthesis and
their impact on the quality of life. Each item consists of a
paired question: the first asks about the extent of a spe-
cific problem during the last 4 weeks and the second
about the impact on quality of life of that specific prob-
lem. Ten items concern problems regardless of prosthetic
use and twenty concern problems in connection with
prosthetic use. Answers are given on a 5-point Likert
scale. No not applicable alternative is available, and if
not relevant, the item could be left out and treated as
missing. A minimum number of 15 paired questions must
698
JRRD, Volume 41, Number 5, 2004
be answered for the calculation of the total Problem
score. A higher score indicates more serious problems.
Global Score (3 Items)
Global health is defined as the perception of function
and problems with the current prosthesis and the percep-
tion of the current overall amputation situation. The score
is a summary of three questions to which answers are
given on a 5-point Likert scale. A Global score of 100
indicates the best possible overall situation as measured
by this instrument.
SF-36-Item Health Survey
The SF-36-Item Health Survey is a validated generic
measure of HRQL assessing the concept in eight separate
scales (Physical Functioning, Role FunctioningPhysi-
cal, Bodily Pain, General Health, Vitality, Social Func-
tioning, Role FunctioningEmotional, and Mental
Health) [29]. The first four scales measure physical
health, and the following four measure psychological
well-being. The results can also be presented in two sum-
mary measures, the Physical Component Score (PCS)
and the Mental Component Score (MCS) [30]. Each
scale, as well as PCS and MCS, can take a value between
0 and 100, where a higher score indicates better health.
The validated Swedish version of the instrument was
used [31].
Measurement Framework
Two perspectives can be applied when one is assess-
ing measurement properties: a psychometric approach
and a clinimetric approach [27]. In psychometrics, a sin-
gle construct is measured with multiple items. Assess-
ments of internal consistency of the instrument are based
on statistical analysis of the relationships between the
items. In clinimetrics, the aim is to measure clinical phe-
nomena that are generally believed to comprise several
unrelated patient characteristics or attributes [32]. In our
study, we were influenced by both approaches, consistent
with the suggestion of Marx et al. that they could be com-
plementary strategies for the development of health meas-
urement scales [32]. Specifically, we began the ongoing
process of validating the Q-TFA by determining its clini-
cal sensibility, criterion validity, test-retest reliability, and
internal consistency.
Clinical sensibility, or face validity, implies that the
instrument is based on clinical knowledge and that the
items included are relevant to the target population [33].
Clinical sensibility is determined qualitatively through
expert opinion and more objectively by examining item
response patterns.
Criterion validity is determined by an analysis of the
extent to which the new tool is related to an existing
gold standard. In the absence of a gold standard, an
already validated instrument measuring the same or simi-
lar constructs could be used. For this study, we chose the
SF-36-Item Health Survey to assess criterion validity of
Q-TFA because of its documented capability to detect
changes in health, its overall widespread use, and its use
with individuals with amputation [2,12,34,35]. The fol-
lowing hypotheses were made:
1. The Prosthetic Use score should show moderate posi-
tive correlations with Physical Functioning and PCS
but no or weak correlations with MCS.
2. The Prosthetic Mobility score should show strong
positive correlations with Physical Functioning and
PCS and no or weak correlations with MCS.
3. The Problem score should be negatively correlated to
all dimensions of SF-36-Item Health Survey, but
those primarily describing physical function should
be more strongly related than the others.
4. The Global score should show positive correlations
with all subscores of SF-36-Item Health Survey.
Reliability was assessed in two ways, test-retest reli-
ability and internal consistency. Test-retest reliability was
assessed by analysis of the differences between the
repeated measurements of the subgroup of 48 partici-
pants living in Vstra Gtaland [36,37].
Statistical Methods
For descriptive purposes, we calculated the mean,
standard deviation (SD), median (Med), and range for
demographic and questionnaire variables. The Mann-
Whitney U-test and Fishers exact test were used to deter-
mine whether there were differences between the retest
subgroup and the remaining study sample. Floor and ceil-
ing effects of each score of the Q-TFA were calculated as
the percentage of subjects achieving the lowest and high-
est score possible. Criterion validity was determined with
Spearmans nonparametric correlation coefficient (rs)
between the subscores of the Q-TFA and the SF-36-Item
Health Survey.
We determined test-retest reliability by descriptive
statistics for each occasion and the differences between
them, calculations of intraindividual SD, measurement
error [38], and intraclass correlation (ICC). We calculated

intraindividual SD by dividing the within-person variance
by the number of participants. ICC is a measure of the
strength of agreement between repeated measurements,
using a one-way analysis of variance model. An ICC of
0.70 or more is recommended for comparisons between
groups and 0.90 for evaluation of individual patients [27].
The Wilcoxon Signed Rank Test was used for hypothesis
testing of the differences between the mean of scores for
occasions 1 and 2. All tests were two-tailed and con-
ducted at a 5-percent significance level.
Internal consistency was analyzed with Cronbachs
alpha. Alpha coefficients higher than 0.70 are generally
recommended to reflect relation between items within a
score [27]. The corrected item total correlation between
each item or subscore and the items overall score was
calculated with Pearsons product-moment correlation (r)
for descriptive purposes [28]. A correlation of r = 0.4 or
higher has been shown to be acceptable [27]. Statistical
calculations were performed with SPSS 10.1 for Win-
dows (SPSS Inc., Chicago, IL).
RESULTS
Participants
A total of 156 of 204 (76%) of the eligible persons
answered the questionnaires. Demographic information
and baseline characteristics of the study population and
the subgroup constituting the test-retest sample (n = 48)
are presented in Table 1. The participants in the retest
subgroup were older (p = 0.011), they had a larger inter-
val since amputation (p = 0.001), and a higher proportion
had a prosthetic limb with vacuum suspension (p = 0.039)
than the remaining study sample (n = 108). No statisti-
cally significant differences between the groups were
found regarding sex (p = 0.36), age at amputation (p =
0.084), presence of other medical problems (p = 0.16),
and reason for amputation divided into vascular or nonva-
scular reasons (p = 1.0). Descriptive statistics as well as
floor and ceiling effects of each score on the Q-TFA are
shown in Table 2.
Clinical Sensibility
Clinical sensibility of the Q-TFA is supported by the
development procedures used, which involved expert
opinions, review of the literature, semistructured inter-
views with experienced prosthetic users, and testing of
the questionnaire on the target population. The content of
699
HAGBERG et al. Questionnaire after transfemoral amputation
the Q-TFA is relevant because very few items were left
unanswered and each item of the questionnaire was
answered with every possible option.
Criterion Validity
Correlations between the scores of the Q-TFA and
scores of the SF-36 are shown in Table 3. As hypothe-
sized, the Prosthetic Use scores were more highly corre-
lated to the physical subscales and the PCS (range of 0.24
to 0.36) than to the mental subscales and the MCS (range
of 0.11 to 0.30). Likewise, the Prosthetic Mobility scores
were generally more highly correlated to the physical sub-
scales and the PCS (range of 0.38 to 0.79) than to the men-
tal subscales and the MCS (range of 0.10 to 0.44). The
Problem score, as hypothesized, was inversely correlated
with all the SF-36 scales (range of 0.30 to 0.68). In
accordance with the final hypothesis, the Global score was
correlated with all the SF-36 scales (range of 0.27 to 0.62).
Test-Retest Reliability
Agreement between test and retest for each subscale
are illustrated with scatter plots in the Figure. Table 4 pre-
sents descriptive measures, intraindividual SD, measure-
ment error, ICC, and hypothesis testing of mean differences
for each subscale. ICC values ranged from 0.89 to 0.94, and
measurement error ranged from 10 to 19 points.
Internal Consistency
The corrected item total correlations ranged from
0.57 to 0.74 in the Mobility score, from 0.39 to 0.83 in
Problem, and from 0.66 to 0.73 in the Global score. Inter-
nal consistency shown by Cronbachs alpha was 0.80 in
Mobility, 0.94 in Problem, and 0.83 in the Global score.
Within the Mobility score, the 12 items of the capability
subscore showed corrected item total correlations
between 0.40 and 0.65 with a total Cronbachs alpha of
0.86; for each of the 2 items of the walking aids subscore,
r was 0.65 with a Cronbachs alpha of 0.78; and for the
5 items constituting the walking habits subscore, the cor-
rected item total correlations ranged from 0.55 to 0.83,
with a Cronbachs alpha of 0.85.
DISCUSSION
This study demonstrates adequate clinical sensibil-
ity, criterion validity, test-retest reliability, and internal
700

JRRD, Volume 41, Number 5, 2004

Table 1.
Demographic and baseline information of study population (N = 156) and subgroup of test-retest sample (n = 48).
Study Population Test-Retest Subgroup
Demographic/Baseline Information
N (%) Mean SD, Med (Range) n (%) Mean SD, Med (Range)
Male 104 (67) 35 (73)
Female 52 (33) 13 (27)
Age 51 13.1, 53 (2070) 55 11.6, 58 (3270)
Years Since Amputation 25 15.6, 24 (256) 32 16.2, 30 (256)
Age at Amputation 26 15.0, 26 (066) 23 15.2, 19.5 (063)
Cause of Amputation
Trauma 86 (55) 30 (63)
Tumor 48 (31) 12 (25)
Vascular 13 (8) 4 (8)
Other 9 (6) 2 (4)
Vacuum Socket (Current Prosthesis) 141 (90) 47 (98)
Other Prosthesis 15 (10) 1 (2)
Presence of Other Medical Problem 71 (46) 26 (54)
SD = standard deviation Med = median

Table 2.
Scores of Questionnaire for Persons with a Transfemoral Amputation (Q-TFA) showing descriptive statistics and percentage of floor and ceiling
effects (N = 156).
Variable Prosthetic Use Prosthetic Mobility Problem Global
Mean 79 67 34 60
Median 90 71 30 58
SD 25 21 20 21
Minimum/Maximum 2/100 3/98 1/84 0/100
% Floor 0 0 0 1
% Ceiling 31 0 0 5
SD = standard deviation

Table 3.
Spearmans correlation coefficient (rs) for scales of Short-Form 36 (SF-36)-Item Health Survey and scores of Questionnaire for Persons with a
Transfemoral Amputation (Q-TFA).
Q-TFA Score
SF-36-Item Health Survey Scale n
Prosthetic Use Prosthetic Mobility Problem Global
Physical Functioning 156 0.36* 0.79* 0.65* 0.59*
Role FunctioningPhysical 154 0.26 0.49* 0.59* 0.53*
Bodily Pain 156 0.24 0.55* 0.62* 0.52*
General Health 155 0.27 0.38* 0.48* 0.50*
Vitality 156 0.16 (NS) 0.20 0.43* 0.42*
Social Functioning 156 0.30* 0.44* 0.61* 0.52*
Role FunctioningEmotional 154 0.11 (NS) 0.32* 0.34* 0.34*
Mental Health 156 0.23 0.22 0.45* 0.40*
Physical Component Score 152 0.34* 0.70* 0.68* 0.62*
Mental Component Score 152 0.19 (NS) 0.10 (NS) 0.30* 0.27
*p < 0.001 p < 0.01 p < 0.05 NS = not significant

Figure 1.
Scatter plots of test-retest for (a) Prosthetic Use, (b) Prosthetic Mobility, (c) Problem, and (d) Global scores of Questionnaire for Persons with
a Transfemoral Amputation (Q-TFA), n = 48.
consistency of the scores of Q-TFA for persons using a
transfemoral socket prosthesis in Sweden.
Q-TFA is primarily intended to be used with the non-
elderly amputee population, among which high prosthetic
use has been reported [5,16,3941]. The age limit of 20 to
70 years in this study was set to follow the current age
criteria for treatment with a bone-anchored prosthesis.
Sixteen percent (n = 25) of the participants were older
than 64 years, and eight percent (n = 13) had an amputa-
tion due to vascular disease. The vascular cases had an
age range between 44 and 70 years (mean 61 years), and
nine of them used a prosthesis daily, showing that some
701
HAGBERG et al. Questionnaire after transfemoral amputation
individuals having a transfemoral amputation for vascular
disease are younger and use their prosthesis extensively.
Extensive problems related to prosthetic use have
been reported, suggesting that it is not sufficient to present
results solely on the amount of prosthetic use if the pur-
pose is to reflect the situation for the individual
[5,12,13,15,16,41,42]. In our development of the Q-TFA,
we tried to capture a broad picture of what it is like to
have to use a transfemoral prosthetic limb, and the sub-
scores concerning prosthetic mobility, problems, and
global health were developed to meet these requirements.
The term mobility has previously been defined as the
702

JRRD, Volume 41, Number 5, 2004

Table 4.
Test-retest reliability of Prosthetic Use, Prosthetic Mobility, Problem, and Global scores when completed on occasions 1 and 2 (n = 48).
Occasion 1 Occasion 2 Difference, Occasion 2 Intraindividual SD p-Value
Score Mean SD, Mean SD, Occasion 1 (Measurement Wilcoxon Signed ICC
Median (Range) Median (Range) Mean SD, Median (Range) Error*) Rank Test
Prosthetic Use 85.0 17.5, 86.7 17.6, 1.7 5.8, 4.2
90.3 90.3 0 (29.0) (12) 0.028 0.94
(32.3 100) (32.7 100)

Prosthetic Mobility 70.7 18.2, 71.0 19.2, 0.4 4.9, 3.5

71.9 75.3 0 (27.8) (10) 0.39 (NS) 0.97
(7.2 96.7) (1.7 98.3)

Problem 28.0 17.7, 25.4 17.7, 2.7 7.8, 5.8

26.5 22.3 1.3 (42.0) (16) 0.026 0.89
(1.3 80.8) (1.3 67.5)

Global 62.7 21.3, 63.4 19.9, 0.7 9.7, 6.8

58.3 62.5 0 (50.0) (19) 0.80 (NS) 0.89
(16.7 100) (8.3 100)
*Measurement error = 1.96 2 intraindividual standard deviation (SD) ICC = intraclass correlation coefficient NS = not significant

capability and performance of moving oneself and chang-
ing and maintaining postures [26]. We added the word-
ing when using the prosthesis. Capability and
performance are two different perspectives of mobility,
reflecting issues of can do in the first and do do in the
second [26,43]. The Mobility score of Q-TFA is intended
to reflect both perspectives. In the subscore of capability,
only those items answered with Yes are counted to
reflect what the individual really is able to perform.
A main concern among persons using a transfemoral
socket prosthesis is socket-related problems. We empha-
sized these problems in the Problem score to make it sensi-
tive to changes for the assessment of individuals being
supplied with a bone-anchored prosthetic limb. The 4-week
interval in the Problem score was chosen to be in line with
the SF-36-Item Health Survey. However, the items con-
cerning different weather conditions (29 and 30 in the
Problem score) and the Walking Habits subscore within the
Mobility score had a longer timeframe due to the changing
seasons in Sweden, which result in different conditions
(Appendix, online version only).
Clinical Sensibility
Determination of clinical sensibility involved qualita-
tive analysis rather than statistical testing. When a new
questionnaire is developed, the item-generation process
should include input from specialists from the area of
interest, a review of existing literature, and interviews with
patients to ensure that the content of the questionnaire cov-
ers relevant and important issues [27]. We considered
these requirements in developing the Q-TFA. The fact that
every individual item included in the scores was answered
with the full range of options that were presented further
supports the relevance of the content toward the target
population. Another concern is the ease of answering the
questionnaire. One indication of ease is a low number of
missing answers. In this study, the number of missing
answers of separate items was, in general, very low, and
enough of the items were answered to be able to calculate
all four scores for each participant. Some individuals,
however, needed additional information on how to answer
the walking habits subscore, and six individuals (4%) did
not answer all five questions of this subscore, which indi-
cates a possible need to change the wording of this specific
question.
Criterion Validity
In this study, we have begun the process of criterion
validation by assessing the relationship between scores of
the SF-36-Item Health Survey and the Q-TFA. Correla-
tion coefficients of r = 0.4 to 0.8 are considered adequate
between abstract constructs such as those in generic
HRQL measures and a new instrument [28]. We achieved
these levels of correlation for most of the hypothesized
relationships, and the associations between the SF-36-
Item Health Survey and Q-TFA were in accordance with

the prior hypotheses made to assure criterion validity in
this study. We had, however, expected a somewhat greater
correlation between the Physical Function and Prosthetic
Use scores (rs = 0.36) (Table 3). The low association sug-
gests that frequent use of a prosthetic limb is not equiva-
lent to high-level prosthetic physical function.
A lower-limb amputation has previously been
reported to influence mainly the physical dimensions of
the SF-36-Item Health Survey [2,12,34]. The scores of
the Q-TFA were also generally more strongly associated
with those dimensions primarily reflecting physical
health. However, the Problem and Global scores were
also associated, to a substantial degree, to scores reflect-
ing physiological well-being (Table 3), indicating that
the Q-TFA also captures a broad range of health con-
cepts. This study confirms adequate criterion validity of
Q-TFA. Additional research to further assess construct
validity, sensibility, and responsiveness of the tool is
ongoing and will be presented in future studies.
Reliability
Along with the second mailing of the Q-TFA, we
included four additional questions to verify if any impor-
tant change in condition had occurred since the first mail-
ing. Nine individuals were excluded from the calculations
of agreement because of such changes (two volume
changes of the residual limb, two sores on the residual
limb, one flu, one bronchitis, one acute lumbago, one
change to a new prosthesis, one other personal problem).
We are, however, not convinced that we really captured a
stable situation in all the included cases. The clinical
experience is that a true stable condition for persons using
a prosthetic limb is rare and that frequent, sometimes
daily, changes in the condition of the residual limb could
appear, and this could have influenced the test-retest
result. The overall agreement between tests reveals that
scores of Q-TFA are reliable (Table 4). The very small but
statistically significant difference of the mean in the Pros-
thetic Use and Problems scores (+1.7, 2.7) is not consid-
ered to have important clinical relevance. What is more
important is the result of the intraindividual SD, the meas-
urement error, and the ICC. The ICC is a measure of the
variance between subjects on repeated measures, while
the intraindividual SD describes the within-person varia-
tion [36]. The interpretation of the measurement error of
scores of Q-TFA in this study reveals that a difference
larger than 12 in the Prosthetic Use score, 10 in the Mobil-
ity score, 16 in the Problem score, and 19 in the Global
703
HAGBERG et al. Questionnaire after transfemoral amputation
score are needed to claim a real difference, above the
noise, on repeated assessments with 95-percent confi-
dence (Table 4). The results demonstrate less agreement
at retest in the Problem and Global scores than in the Pros-
thetic Use and Mobility scores. Hays et al. claim that the
well-being part of HRQL is more subjective than the func-
tioning component [44]. The same circumstances could
explain the differences of agreement regarding issues of
prosthetic use and mobility on one hand and the more sub-
jective matters of problems and global health on the other.
A more subjective item could be expected to depend more
on the mood of the person answering the questionnaire on
that specific day, leading to larger variation at retest.
In concordance with the clinimetric approach
[27,33], we tried to include items that could be expected
to be important to the target population, regardless of
their influence on the consistency of each score. Never-
theless, we were also interested in assessing whether the
scales exhibited sufficient internal consistency. The
results revealed good internal consistency with no alpha
coefficient below 0.7. Furthermore, only one item was
below the recommended r = 0.4 when calculations of the
corrected item-total correlation were conducted.
Study Limitations
A self-report instrument has the advantages of
reflecting the subjective experience of each participant
and being easy and inexpensive to administer. The limita-
tion is that misunderstandings of answers to the question-
naire are difficult to control. However, because our
primary goal was to describe the current situation from
the view of the patient, with minimal influence from an
investigator, we chose a self-report tool.
The method by which participants were recruited to
the study might be another source of error. Presently, no
national registration exists of persons with a lower-limb
amputation in Sweden, and potential participants were
identified through amputees associations and selected
orthopedic workshops. Thus, some potential participants
may not have been reached. The number of participants
in the study is, however, satisfactory with regard to the
inclusion criteria used. Only adults, 20 to 70 years of age,
who had been amputated at least 2 years earlier and who
currently were using a prosthetic limb, were included.
The age limit was set to reflect the population that could
be considered potential candidates to be treated with a
bone-anchored prosthesis. To reach persons that could be
704
JRRD, Volume 41, Number 5, 2004
considered established prosthetic users, we chose a mini-
mum 2-year interval since the amputation.
CONCLUSIONS
We developed a new self-report questionnaire, the Q-
TFA, to assess nonelderly persons having a transfemoral
amputation and using a prosthesis. This study of 156
individuals with transfemoral amputation living in Swe-
den documents adequate clinical sensibility, criterion
validity, test-retest reliability, and internal consistency for
the Q-TFA for persons using a transfemoral socket pros-
thesis. Further assessment of the construct validity, sensi-
tivity, and responsiveness of the Q-TFA is needed.
REFERENCES
1. Pell JP, Donnan PT, Fowkes FG, Ruckley CV. Quality of
life following lower limb amputation for peripheral arterial
disease. Eur J Vasc Surg. 1993;7(4):44851.
2. Pezzin LE, Dillingham TR, MacKenzie EJ. Rehabilitation
and the long-term outcomes of persons with trauma-related
amputations. Arch Phys Med Rehabil. 2000;81(3):292300.
3. Burger H, Marincek C. The life style of young persons
after lower limb amputation caused by injury. Prosthet
Orthot Int. 1997;21(1):3539.
4. van der Schans CP, Geertzen JH, Schoppen T, Dijkstra PU.
Phantom pain and health-related quality of life in lower limb
amputees. J Pain Symptom Manage. 2002;24(4):42936.
5. Schoppen T. Functional outcome after a lower limb amputa-
tion. Groningen: University of Groningen, Netherlands; 2002.
6. Ebskov LB. Level of lower limb amputation in relation to
etiology: an epidemiological study. Prosthet Orthot Int.
1992;16(3):16367.
7. Pohjolainen T, Alaranta H. Ten-year survival of Finnish
lower limb amputees. Prosthet Orthot Int. 1998;22(1):1016.
8. Pernot HF, de Witte LP, Lindeman E, Cluitmans J. Daily
functioning of the lower extremity amputee: an overview
of the literature. Clin Rehabil. 1997;11(2):93106.
9. Holden JM, Fernie GR. Extent of artificial limb use follow-
ing rehabilitation. J Orthop Res. 1987;5(4):56268.
10. Gauthier-Gagnon C, Grise MC, Potvin D. Enabling factors
related to prosthetic use by people with transtibial and
transfemoral amputation. Arch Phys Med Rehabil. 1999;
80(6):70613.
11. Gautier-Gagnon C, Grise MC, Potvin D. Predisposing fac-
tors related to prosthetic use by people with a transtibial
and transfemoral amputation. J Prosthet Orthot. 1998;10:
99109.
12. Legro MW, Reiber G, del Aguila M, Ajax MJ, Boone DA,
Larsen JA, Smith DG, Sangeorzan B. Issues of importance
reported by persons with lower limb amputations and pros-
theses. J Rehabil Res Dev. 1999;36(3):15563.
13. Matsen SL, Malchow D, Matsen FA 3rd. Correlations with
patients perspectives of the result of lower-extremity
amputation. J Bone Joint Surg Am. 2000;82-A(8):108995.
14. Hagberg K, Brnemark R. Consequences of nonvascular
trans-femoral amputation: a survey of quality of life, pros-
thetic use and problems. Prosthet Orthot Int. 2001;25(3):
18694.
15. Gallagher P, Maclachlan M. Adjustment to an artificial
limb: a qualitative perspective. J Health Psychol. 2001;
6(1):85100.
16. Dillingham TR, Pezzin LE, MacKenzie EJ, Burgess AR.
Use and satisfaction with prosthetic devices among persons
with trauma-related amputations: a long-term outcome
study. Am J Phys Med Rehabil. 2001;80(8):56371.
17. Dillingham TR, Pezzin LE, MacKenzie EJ. Limb amputa-
tion and limb deficiency: epidemiology and recent trends in
the United States. South Med J. 2002;95(8):87583.
18. Ebskov LB. Trauma-related major lower limb amputations:
an epidemiologic study. J Trauma. 1994;36(6):77883.
19. Brnemark R, Brnemark PI, Rydevik B, Myers RR.
Osseointegration in skeletal reconstruction and rehabilita-
tion: a review. J Rehabil Res Dev. 2001;38(2):17581.
20. Grise MC, Gauthier-Gagnon C, Martineau GG. Prosthetic
profile of people with lower extremity amputation: concep-
tion and design of a follow-up questionnaire. Arch Phys
Med Rehabil. 1993;74(8):86270.
21. Day HJ. The assessment and description of amputee activ-
ity. Prosthet Orthot Int. 1981;5(1):2328.
22. Miller WC, Deathe AB, Speechley M. Lower extremity
prosthetic mobility: a comparison of 3 self-report scales.
Arch Phys Med Rehabil. 2001;82(10):143240.
23. Legro MW, Reiber GD, Smith DG, del Aguila M, Larsen J,
Boone D. Prosthesis evaluation questionnaire for persons
with lower limb amputations: assessing prosthesis-related
quality of life. Arch Phys Med Rehabil. 1998;79(8):93138.
24. Cella D, Nowinski CJ. Measuring quality of life in chronic
illness: the functional assessment of chronic illness therapy
measurement system. Arch Phys Med Rehabil. 2002;83(12
Suppl 2):S1017.
25. Patrick DL, Chiang YP. Measurement of health outcomes in
treatment effectiveness evaluations: conceptual and method-
ological challenges. Med Care. 2000;38 Suppl 9:II1425.
26. Bussmann JB, Stam HJ. Techniques for measurement and
assessment of mobility in rehabilitation: a theoretical
approach. Clin Rehabil. 1998;12(6):45564.
27. Fayers PM, David M. Quality of life: assessment, analysis
and interpretation. Chichester: John Wiley & Sons, Ltd;
2000.

28. Streiner DL, Norman GR. Health measurements scales: a
practical guide to their development and use. 2nd ed. New
York: Oxford University Press; 1995.
29. Ware JE Jr, Sherbourne CD. The MOS 36-item short-form
health survey (SF-36). I. Conceptual framework and item
selection. Med Care. 1992;30(6):47383.
30. Ware JE Jr, Kosinski M, Bayliss MS, McHorney CA, Rog-
ers WH, Raczek A. Comparison of methods for the scoring
and statistical analysis of SF-36 health profile and sum-
mary measures: summary of results from the Medical Out-
comes Study. Med Care. 1995;33 Suppl 4:AS26479.
31. Sullivan M, Karlsson J, Ware JE. SF-36 Hlsoenkt:
Svensk manual och tolkningsguide [Swedish manual and
interpretation guide]. Gothenburg: Sahlgrenska University
Hospital; 1994.
32. Marx RG, Bombardier C, Hogg-Johnson S, Wright JG. Clin-
imetric and psychometric strategies for development of a
health measurement scale. J Clin Epidemiol. 1999;52(2):
10511.
33. Feinstein AR. Clinimetrics. New Haven and London: Yale
University Press; 1987.
34. Smith DG, Horn P, Malchow D, Boone DA, Reiber GE,
Hansen ST Jr. Prosthetic history, prosthetic charges, and
functional outcome of the isolated, traumatic below-knee
amputee. J Trauma. 1995;38(1):4447.
35. Beaton DE, Hogg-Johnson S, Bombardier C. Evaluating
changes in health status: reliability and responsiveness of
five generic health status measures in workers with muscu-
loskeletal disorders. J Clin Epidemiol. 1997;50(1):7993.
36. Rankin G, Stokes M. Reliability of assessment tools in reha-
bilitation: an illustration of appropriate statistical analyses.
Clin Rehabil. 1998;12(3):18799.
705
HAGBERG et al. Questionnaire after transfemoral amputation
37. Bland JM, Altman DG. Statistical methods for assessing
agreement between two methods of clinical measurement.
Lancet. 1986;1(8476):30710.
38. Bland JM, Altman DG. Measurement error. BMJ. 1996;
313(7059):744.
39. Burger H, Marincek C, Isakov E. Mobility of persons after
traumatic lower limb amputation. Disabil Rehabil. 1997;
19(7):27277.
40. Hoffman RD, Saltzman CL, Buckwalter JA. Outcome of
lower extremity malignancy survivors treated with trans-
femoral amputation. Arch Phys Med Rehabil. 2002;83(2):
17782.
41. Walker CR, Ingram RR, Hullin MG, McCreath SW. Lower
limb amputation following injury: a survey of long-term
functional outcome. Injury. 1994;25(6):38792.
42. Gallagher P, Allen D, Maclachlan M. Phantom limb pain
and residual limb pain following lower limb amputation: a
descriptive analysis. Disabil Rehabil. 2001;23(12):52230.
43. Young NL, Williams JI, Yoshida KK, Bombardier C,
Wright JG. The context of measuring disability: does it
matter whether capability or performance is measured?
J Clin Epidemiol. 1996;49(10):1097101.
44. Hays RD, Hahn H, Marshall G. Use of the SF-36 and other
health-related quality of life measures to assess persons
with disabilities. Arch Phys Med Rehabil. 2002;83(12
Suppl 2):49.
Submitted for publication November 13, 2003. Accepted
in revised form April 14, 2004.